Glucocorticoids constitute a class of molecules widely used in animal husbandry. Some of these compounds are licensed for veterinary practices while their use for growth promoting purposes is prohibited within the European Union. In order to ensure the respect of the legislation and consumers safety, several methodologies have been proposed to monitor these substances in various products, including edible matrices for which a regulatory limit has been set up (MRL). An extended range of targeted analytes together with reduced time of analysis and cost are however still current challenges regularly revisited according to the continuous technological improvements. In this context, the aim of the present study was to develop and implement a new fast and multi-residue method based on UHPLC-MS/MS for the determination of twenty glucocorticoids in bovine milk, included the screening of the three regulated MRL compounds (dexamethasone, betamethasone and prednisolone). This validated method authorises such multi-analyte measurement within a 10 min runtime while the signal specificity is ensured through the SRM acquisition mode. Decision limits and detection capabilities were calculated in the range of 0.001-0.363 mu g L-1, which allows a very efficient control at low trace level for a potential illegal use of these substances. The performances obtained in terms of application range, selectivity and sensitivity were found to be significantly improved in comparison to other reported approaches either for screening or confirmation purposes: regarding linearity, correlation coefficients were above 0.98 within the range of 0.01-5.0 mu g L-1, repeatability and reproducibility parameters ranged from 1 to 30% with the maximum relative standard deviation (RSD) observed for cortisone (30.1%). Stability of the stock solutions and minor changes in the standard operating procedure have been included for the determination of ruggedness of the method. Identification was systematically ensured according to 4 identification points, RSD of transitions ratio (T2/T1) ranged from 3.2% and 19.3% and the RSD of the retention time was lower than 0.25%. (C) 2013 Elsevier B.V. All rights reserved.